Volatility Forecasting using Deep Learning-Bidirectional LSTM for Financial Stock Market

Abstract

In this paper, we have proposed a novel BiLSTM technique for stock price forecasting. BiLSTM technique is used for the prediction of a stock price also it solves the problem of fixed sequence to sequence prediction with the use of data preprocessing, normalization, training and testing. The tests performed using deep learning techniques to forecast volatility. The methodology consists of two datasets AAPL & S&P 500 that are part of yahoo finance.   The findings indicate that the methods of BiLSTM have achieved higher accuracy than existing LSTM-RNN. These findings demonstrated a high level of reliability and predictability of the proposed model.